DISCUSSION
Fibrinolysis is a physiologic mechanism by which clots are degraded to maintain the patency of microvasculature. Pathologic over- or under-activation of fibrinolysis has significant clinical implications regarding optimal treatment interventions and patient outcomes in trauma patients22-28 . This study is the first to categorize these fibrinolysis phenotypes in cardiac surgery patients. The data suggest that indiscriminate AF administration may have deleterious effects unless clinically indicated, but the sample size is too small to make definitive recommendations.
Hyper-, physiologic, and hypo-fibrinolytic phenotypes are well established among trauma patients. In theory, AF should have its most favorable effect in those with hyperfibrinolysis. In contrast, hypercoagulability associated with fibrinolytic shutdown is reported to occur in 46% to over 64% of trauma patients22-32. On this basis alone, it is reasonable to propose that all patients likely do not benefit from TXA11. Despite the high prevalence, it remains controversial if fibrinolytic shutdown is a physiological or maladaptive response to traumatic injury33, 34. Our study demonstrates a distribution of fibrinolytic phenotypes in cardiac surgery patients that is distinctly different than the established distributions among trauma patients. In particular, our data shows that cardiac surgery patients exhibited a rate of fibrinolytic shutdown of 32% compared to the reported range for trauma patients of 46-64%.
There are multiple proposed mechanisms by which fibrinolytic shutdown occurs, including: i) tissue plasminogen activator (t-PA) inhibition, ii) inadequate t-PA release in response to injury, iii) fibrinolytic resistance via cell free DNA, and iv) elevated plasmin to antiplasmin ratio26, 30,32, 35, 36. In vivo studies in animal models have demonstrated that tPA release is predominantly driven by shock rather than by tissue injury and that altered clot formation has no correlation with altered clot degradation22,36, 37. Subsequent studies have yet to identify a correlation between injury severity score or injury mechanism and fibrinolytic phenotype36, 38. Nevertheless, changes in the fibrinolytic system after tissue injury have significant treatment implications regarding the treatment decision to administer or withhold AF therapy.
Although distinct fibrinolytic phenotypes are well-established in trauma patients, equivalent phenotypes are yet to be fully characterized in other surgical fields. Many studies have investigated the efficacy of empiric AF treatment across surgical groups; however, there is limited discussion of the prevalence or influence of distinct fibrinolytic phenotypes on treatment responsiveness and outcomes. In particular, there may be an increased potential for preventable harm if AF is administered to a patient who is already hypofibrinolytic (i.e. fibrinolysis shutdown=hypercoagulable), or if withheld from a patient who is hyperfibrinolytic (i.e. hypocoagulable).
In the United States, patients undergoing cardiac surgery demonstrate particularly high blood transfusion rates. In 2010, 34% of cardiac surgery patients received a perioperative transfusion despite the implementation of blood conservation guidelines in 200739-41. Thus, AF is often used in this population to minimize perioperative blood loss 42. It is well established that AF therapy reduces bleeding and allogeneic transfusion requirements29. Myles et al studied the effectiveness of TXA versus placebo in a randomized control study of 4631 coronary-artery surgical patients and demonstrated that death or thrombotic complication occurred in 18.1% of placebo patients versus 16.7% of TXA patients43. Transfusion requirements were reduced from 7,994 total units in placebo to 4,331 in the TXA group, with seizures reported in 0.1% of placebo versus 0.7% of TXA patients43. A recent metanalysis by Alaifan et al reported that TXA reduced bleeding in cardiac surgery, but surprisingly did not significantly impact overall mortality49. A randomized control trial by Leff at el comparing TXA and ACA in 114 cardiac surgery patients demonstrated that ACA was associated with less transfusions than TXA; though, both TXA and ACA significantly reduced perioperative bleeding and transfusions with no increase in adverse events45. The efficacy of AF in reducing perioperative bleeding and transfusion requirements has been reported by several others46-48.
Our data demonstrated significant outcome differences in those who received AF versus those who did not. We show that cardiac surgery patients who received AF had significantly higher rate of all cause morbidity (n = 19, 51%) versus those who did not (n = 10, 25%, P = 0.017). Patients who received AF also had more days with a chest tube (P = 0.037) and an average of 559mL more output from the chest tube compared to patients who did not receive TXA (P = 0.075). Unfortunately, the sample size was not large enough to determine if these outcome differences were related to the fibrinolytic phenotype.
There was no obvious effect of age, gender, race, or ethnicity on the distribution of fibrinolysis phenotypes. There was a near equal distribution of physiologic, hyper-, and hypo-fibrinolytic phenotypes between patients who received AF (46% vs 30% vs 24%) and did not receive AF (45% vs 33% vs 22%, P = 0.962). This suggests that a patient’s fibrinolytic phenotype was not a contributing factor in the decision to administer or withhold AF.
The use of AF in other surgical populations has generally been efficacious; though, there is significant heterogeneity across specialties. For instance, patients with hepatic dysfunction are at a markedly increased risk of excessive bleeding or coagulopathy during liver surgery or transplantation due to altered hepatic production of essential clotting factors49. Previously, AF were administered empirically in these patients; however, empiric therapy has since been questioned due an observed increase in rates of coagulopathy, venous thromboembolism, and mortality49-51. In current practice, the use of AF during hepatic operations is variable due to the considerable physiologic changes that occur pre- and post-liver transplantation. Accordingly, most liver transplant centers use TEG or viscoelastic monitoring with rotational thromboelastometry to guide AF administration if a patient exhibits hyperfibrinolysis52-54. Similarly, it has been established that only some trauma patients benefit from empiric AF therapy. Thus, the use of AF in trauma is generally informed and guided by TEG or comparable monitoring techniques.
The practical significance of fibrinolytic phenotype and therapeutic response has only recently been recognized. Even among the fields of liver and trauma surgery, where fibrinolytic variability has been consistently reported, there is still limited discussion of the relationship between AF and clinical outcomes across distinct fibrinolytic phenotypes.
Ultimately, further investigation is needed to assess the role of fibrinolytic phenotypes in modulating AF responsiveness and clinical outcomes, especially because AF therapy has a demonstrated utility in cardiac surgery. Ultimately, TEG -guided AF therapy might assure administration only when clinically indicated..
Our study is limited by its small sample size and retrospective methodology, which precluded subgroup analysis of AF outcomes across fibrinolytic phenotypes. Our small study size also hindered statistical significance of some outcome variables; though, we still observed clinically significant differences among patients who received AF, including an average of 325mL more EBL (P = 0.127) and 2 units more of RBCs transfused (P = 0.152) compared to patients who did not receive AF. Considering that fibrinolytic phenotypes and most demographic factors were relatively equal between patients who did and did not receive AF, these observed differences following AF may suggest a contributory role of fibrinolytic phenotype on patient outcomes. Therefore, a better understanding of fibrinolytic phenotypes may be especially significant in guiding the decision to administer or withhold AF in surgical patients.
In summary, this study is the first to describe three distinct fibrinolytic phenotypes in cardiac surgery patients. This distribution is different than the established distribution of fibrinolytic phenotypes in trauma patients. In most surgical specialties, intra-operative AF safely and effectively reduces bleeding complications, but there is on strong evidence from trauma patients that fibrinolytic phenotypes in modulating treatment responses to AF. In context with these current findings, we hope to open the dialog on whether it is safe to administer AFs to cardiac surgery patients who are normo- or hypofibrinolytic. Further studies are clearly indicated in cardiac surgery patients to investigate the clinical utility of pre- and post-operative TEG data to discern fibrinolytic phenotypes and guide AF use.